The imprecision of our estimate (ie, 95% CI –2 to 15) was greater than expected and greater than a comparable study upon which we based our power calculations (95% CI 4 to 7, Bakhtiary and Fatemy 2008). There are differences between our trial and that of Bakhtiary and Fatemy which may explain these differences. Our trial recruited people with obvious weakness, and either spasticty or reduced inhibitors extensibility of the long finger flexor muscles after an acquired brain injury regardless of anti-spasticity medication, whereas Bakhtiary and Fatemy recruited patients with spasticity after stroke who were not receiving anti-spasticity medication. It is possible that the two

groups of patients Epacadostat ic50 respond differently to electrical stimulation. The electrical stimulation protocols were also different. In our trial, electrical stimulation was applied at the maximal tolerable intensity for 1 hour a day whereas Bakhtiary and Fatemy applied supramaximal levels of electrical stimulation (ie, the intensity was set at 25% over the intensity needed to produce a maximum contraction) for 9 minutes a day. It is not clear how participants tolerated such high doses of electrical stimulation. Another difference is that in our trial electrical stimulation was applied with the wrist held in an extended position in order to optimise any beneficial stretching

and strengthening effects. In contrast, Bakhtiary and Fatemy applied electrical stimulation with the ankle unsupported (and presumably in a plantarflexed position). We are not sure if enough any of these differences between the two trials are important. There are buy GS-7340 other factors that may explain the imprecision of our estimate of treatment effectiveness. First, there was considerable variability in the participants’ age, length of time post-injury, and degree of spasticity,

weakness, motor control, and hand contracture. These factors may vary the way participants responded to the intervention. Second, some participants in our study had difficulty relaxing during measures of passive wrist extension because of pain. Although any inadvertent muscle activity was unlikely to bias the results systematically, it may have added noise to the data leading to an imprecise estimate (ie, wide 95% CI). Perhaps there are sub-groups of participants who respond more favourably to electrical stimulation than others. For instance, initial strength may be an important determinant of the effectiveness of electrical stimulation. There is growing evidence to suggest that electrical stimulation may be more effective for increasing strength when combined with voluntary movements or functional activity (Alon et al 2008, Bolton et al 2004, Chan et al 2009, de Kroon et al 2002, Ng and Hui-Chan 2007). It is possible that people with some strength in their wrist or finger extensor muscles benefit more from electrical stimulation than those without any strength.